The G-CSF Receptor Requires Gab2-Mediated Recruitment of the Tyrosine Phosphatase Shp2 to Promote Lyn-Dependent Proliferation.

Abstract 1555

Hematopoietic cytokine receptors, such as the G-CSFR, use Janus and Src kinases to transduce their signal. Less well known is how the receptors activate these cytosolic protein tyrosine kinases. The phosphorylation/dephosphorylation of inhibitory and stimulatory tyrosine residues of the Src kinases are critical regulatory steps. For Lyn, phosphorylation of Y507 inhibits its activity, whereas phosphorylation of Y396 promotes it. We hypothesized that the tyrosine phosphatase SHP2 is activated by G-CSFR signaling, resulting in dephosphorylation at the negative regulatory site Lyn Y507, and that the adaptor protein Gab2 directs SHP2 effects on phospho-Lyn Y507. To address this hypothesis, we established mouse IL-3-dependent Ba/F3 cells which express the G-CSFR (Ba/F3GR cells). (1) After G-CSF stimulation, phosphorylation status of Lyn (Y507, Y396) was determined by immunoblotting, and protein-protein binding between Gab2-SHP2 and Gab2-Lyn was assessed by the immunoprecipitation and the GST-pull down assay. (2) Ba/F3GR cells were transfected with SHP2, Gab2, or the phosphorylation defective Gab2 mutant (Y614F, Y643F), and phosphorylation status of Lyn was determined. (3) G-CSF-dependent proliferation and colony formation was determined after knockdown of Gab2 or SHP2. After stimulation of Ba/F3GR cells with G-CSF, the inhibitory phosphorylation site Lyn Y507 was dephosphorylated, and the activating site Lyn Y396 was phosphorylated, due to the autophosphorylation. This was observed as a spike at 5 min of the G-CSF treatment and returned gradually to the basal state. When SHP2-deficient cells were treated with G-CSF, phospho-Lyn Y507 was not dephosphorylated, whereas a constitutively active SHP2 mutant E76A dephosphorylated Lyn Y507. In addition, when SHP2 was overexpressed in Ba/F3GR cells, prolonged dephosphorylation of Lyn Y507 was observed. These data suggest that SHP2 is involved in the G-CSF-induced activation of Lyn by the dephosphorylation of Lyn Y507. Co-immunoprecipitation studies revealed that an adapter protein Gab2 binds to Lyn, and this binding is abrogated by the G-CSF treatment in a time course similar to the dephosphorylation of phospho-Lyn Y507. Gab2 has two tyrosine residues (Y614, Y643), which, when phosphorylated, provide binding sites for the SH2-domain of SHP2. Engagement of SHP2 through its SH2-phosphotyrosine interaction leads to a conformational change and activates SHP2. We previously reported that G-CSFR signaling phosphorylates Gab2 (Blood 103: 3305, 2004). As expected, wild-type Gab2 co-immunoprecipitated with SHP2, whereas Gab2 double mutant (Y614F, Y643F) did not. When we transfected Ba/F3GR cells with the Gab2 double mutant, G-CSFR-induced dephosphorylation of Lyn Y507 was abrogated. These findings were confirmed by the knockdown experiments using Gab2 siRNA. To examine the biological consequence, G-CSF-dependent proliferation of Ba/F3GR cells was determined. After the knockdown of SHP2 and Gab2, cell proliferation was inhibited significantly (cell number on day 3 was 21% in Ba/F3GR-SHP2 shRNA, and 43% in Ba/F3GR-Gab2 shRNA, compared to Ba/F3GR-control shRNA). Furthermore, G-CSF-induced CFU-G colony formation was inhibited in bone marrow cells from Gab2^-/-mice compared to wildtype mice (Gab2^-/- 17.7 ± 0.9 vs. wildtype 25.7 ± 3.8). In summary, we found that treatment with G-CSF results in the dephosphorylation of the negative regulatory site Lyn Y507 and activation of Lyn, and that the binding of Gab2 to SHP2 was required for G-CSF-dependent cell proliferation. These data support a mechanistic model for G-CSF-induced proliferation that requires the activation of the Src kinase Lyn through recruitment of tyrosine phosphatase Shp2 via the adaptor protein Gab2.